Data from Summative Orthogonal Layout Study

Conducted by Steve Kieffer, Tim Dwyer, Kim Marriott, and Michael Wybrow at Monash University, Caulfield, Victoria, Australia

Contents

Top

Part 1 -- Small Graph Triples

89 respondents (76 for graphs 1, 3, and 8)

Blue column = 2*gold + 1*silver + 0*bronze

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 1 2 3 4 5 6 7 8

HOLA

yFiles

Human

Graph 1

HOLA

yFiles

Human

Graph 2

HOLA

yFiles

Human

Graph 3

HOLA

yFiles

Human

Graph 4

HOLA

yFiles

Human

Graph 5

HOLA

yFiles

Human

Graph 6

HOLA

yFiles

Human

Graph 7

HOLA

yFiles

Human

Graph 8

HOLA

yFiles

Human

Part 2 -- Paths

84 respondents

60-Sp

90-Sp

90-Ds

120-Sp

SBGN

Metro

HOLA

yFiles

Paths for 60-Sp

HOLA

Avg time: 12.78s

73 -- avg. 12.29s

1_6_30

6 -- avg. 15.93s

1_2_4

1 -- avg. 50.75s

21

1 -- avg. 10.03s

1

1 -- avg. 7.78s

1_30

1 -- avg. 9.78s

1_4

1 -- avg. 2.63s

14_24

yFiles

Avg time: 11.07s

43 -- avg. 11.03s

1_2_4

39 -- avg. 10.82s

1_6_30

1 -- avg. 5.74s

1

1 -- avg. 27.84s

6_9_10_11_21_22_23_27_30_31_55

Paths for 90-Sp

HOLA

Avg time: 9.10s

42 -- avg. 9.34s

2_40_79

34 -- avg. 8.86s

2_20_40

6 -- avg. 9.20s

2_5_9_20

1 -- avg. 4.42s

2

1 -- avg. 10.90s

2_79

yFiles

Avg time: 14.76s

36 -- avg. 14.81s

2_40_79

28 -- avg. 14.14s

2_5_9_20

12 -- avg. 15.98s

2_20_40

5 -- avg. 9.79s

2_20

1 -- avg. 47.99s

1_4_8_17_34_40_70_78_79

1 -- avg. 14.43s

40_79

1 -- avg. 7.50s

2

Paths for 90-Ds

HOLA

Avg time: 17.15s

49 -- avg. 16.78s

2_3_9

14 -- avg. 12.03s

1_45_64

6 -- avg. 14.80s

4_25_34_45_64_69

2 -- avg. 26.60s

2_10_26_45

1 -- avg. 9.76s

2_3

1 -- avg. 32.80s

2_3_17_47

1 -- avg. 12.50s

2_9

1 -- avg. 21.91s

2_3_8

1 -- avg. 6.31s

45

1 -- avg. 14.99s

1_17_27_47_84

1 -- avg. 26.97s

3_9

1 -- avg. 38.65s

2_15_44_45_64

1 -- avg. 50.59s

2_3_9_42_47_50_66

1 -- avg. 8.93s

1_17_20_27_47_84

1 -- avg. 30.47s

2_9_15_42_50_66_87

1 -- avg. 9.25s

1_20_84

1 -- avg. 45.00s

2_3_10_26_45

yFiles

Avg time: 33.62s

19 -- avg. 26.61s

1_17_20_27_47_84

11 -- avg. 27.48s

9_11_14_28_72

8 -- avg. 34.22s

2_3_17_47

8 -- avg. 24.47s

1_5_11_14_17_20_27_28_47_84

5 -- avg. 43.37s

2_3_9

5 -- avg. 20.02s

11_14_17_28_47_72

3 -- avg. 19.85s

1_5_11_14_17_20_27_28_35_47_84

2 -- avg. 119.25s

2_10_26_45

2 -- avg. 39.47s

2_15_44_45_64

2 -- avg. 20.50s

44_45_64

2 -- avg. 58.79s

11_14_26_28_45

1 -- avg. 29.81s

3_17_47

1 -- avg. 72.76s

1_17_27_47_84

1 -- avg. 76.80s

1_9_11_14_28_72

1 -- avg. 34.14s

1_5_11_13_14_16_28_44_45_56_64_89

1 -- avg. 44.55s

3_6_9_11_14_28_36_62_70_72

1 -- avg. 19.30s

7_44_45_64_75

1 -- avg. 67.87s

26

1 -- avg. 32.59s

1_5_11_17_20_27_28_35_47_84

1 -- avg. 12.25s

7_44_45_51_64_75

1 -- avg. 42.89s

2_15_17_27_44_47_56_89

1 -- avg. 8.58s

9_28

1 -- avg. 36.39s

1_17_20_27_47

1 -- avg. 88.45s

2_9_15_42_50_66_87

1 -- avg. 6.56s

11

1 -- avg. 29.30s

4_25_34_45_69

1 -- avg. 63.91s

1_5_9_28_72

1 -- avg. 28.31s

2_3_9_17

Paths for 120-Sp

HOLA

Avg time: 14.23s

68 -- avg. 14.52s

0_1_7_11

11 -- avg. 13.69s

7_11_15

2 -- avg. 14.25s

4_6_8_11_39_43_49_53

1 -- avg. 13.56s

1_7_11_15

1 -- avg. 11.14s

7_11_15_17_112

1 -- avg. 4.75s

1

yFiles

Avg time: 27.72s

29 -- avg. 31.35s

7_11_15_17_112

25 -- avg. 22.15s

4_6_8_11_39_43_49_53

15 -- avg. 30.85s

0_1_7_11

2 -- avg. 14.68s

4_6_8_11_39_43

2 -- avg. 27.94s

4_6_11_39_43_49_53

1 -- avg. 25.99s

4_6_8_11_43_49_53

1 -- avg. 33.46s

11_15_17_112

1 -- avg. 42.02s

0_1_3_7_10_11_13

1 -- avg. 54.50s

7_11_17_112

1 -- avg. 14.24s

4_6_8_39_43_49_53

1 -- avg. 5.86s

4

1 -- avg. 34.04s

0_1_3_5_7_11_37

1 -- avg. 12.53s

4_8_39_43_49_53

1 -- avg. 25.32s

0_3_4_5_6_7_8_11_13_37_51_114_115

1 -- avg. 18.36s

4_5_8_11_39_43_49_53

1 -- avg. 51.20s

0_1_3_7_11_37

Paths for SBGN

HOLA

Avg time: 10.99s

77 -- avg. 11.21s

60_81_109_118

4 -- avg. 8.85s

60_81_118

1 -- avg. 6.20s

81_109_118

1 -- avg. 13.45s

60_109_118

1 -- avg. 4.68s

118

yFiles

Avg time: 18.18s

73 -- avg. 17.53s

60_81_109_118

3 -- avg. 26.13s

38_69_70_71_72_81_91_106_109_117_118_119

1 -- avg. 38.18s

60_81_109_117_118

1 -- avg. 17.75s

38_69_71_81_91_106_109_118

1 -- avg. 38.42s

7_22_38_56_69_70_71_72_81_91_106_109_117_119_160

1 -- avg. 9.46s

118

1 -- avg. 12.26s

81_109_118

1 -- avg. 18.31s

38_69_70_71_72_81_91_106_109_117_118

1 -- avg. 24.13s

60_109_118

1 -- avg. 11.08s

60_81

Paths for Metro

HOLA

Avg time: 8.10s

80 -- avg. 8.09s

55_65_142

2 -- avg. 8.99s

55_65_110_142

1 -- avg. 4.64s

55

1 -- avg. 10.71s

55_65_107

yFiles

Avg time: 58.58s

40 -- avg. 39.54s

55_65_142

18 -- avg. 51.18s

55_65_110_142

3 -- avg. 141.20s

3_7_10_11_15_22_24_40_44_53_67_78_81_87_90_105_118_119_122_123_140_141_143_145_165_171

1 -- avg. 84.78s

3_8_11_13_14_19_28_29_33_40_46_53_56_64_74_79_81_84_87_104_105_111_112_118_122_123_124_128_135_140_141_143_145_147_152_153_158_167

1 -- avg. 134.20s

15_19_23_28_33_38_46_54_79_93_107_110_119_125_135_142_153_163_165

1 -- avg. 89.45s

7_55_65_110_119_142_171

1 -- avg. 54.02s

9_15_28_32_33_38_41_48_52_93_107_110_125_135_142_158_163

1 -- avg. 8.46s

110

1 -- avg. 55.04s

11_19_23_38_54_107_110_125_142_163

1 -- avg. 16.01s

107_142

1 -- avg. 114.93s

2_3_11_19_28_30_31_33_40_57_66_72_79_81_83_87_95_105_106_108_115_118_120_122_123_132_135_138_141_143_148_150_151_153_155_156_157_158_160_168_170

1 -- avg. 119.02s

29_55_56_148_171

1 -- avg. 40.92s

38_48_93_107_125_142_163

1 -- avg. 150.89s

3_10_15_22_24_29_40_44_46_53_56_64_67_78_81_87_90_105_110_118_122_140_141_143_145_165

1 -- avg. 133.84s

11_123

1 -- avg. 63.54s

7_110_142_171

1 -- avg. 38.53s

11_19_23_28_33_38_54_79_93_107_123_125_135_142_153_163

1 -- avg. 102.23s

3_7_10_15_22_24_40_44_53_67_78_81_87_90_105_110_118_119_122_140_141_143_145_165_171

1 -- avg. 41.73s

5_36_38_54_58_60_93_94_107_125_142_163_172

1 -- avg. 130.36s

7_8_10_11_13_14_15_22_24_29_44_46_56_64_67_74_90_104_111_112_119_123_124_128_145_147_152_167_171

1 -- avg. 106.40s

3_15_40_53_81_87_105_110_122_140_141_143_145

1 -- avg. 60.52s

55

1 -- avg. 89.06s

3_10_11_19_22_24_28_29_33_40_44_53_56_64_67_78_79_81_84_87_90_105_118_119_122_123_135_140_141_143_145_153_158_165

1 -- avg. 205.07s

7_55_64_65_110_112_171

1 -- avg. 91.22s

28_29_55_56_110_148_171

1 -- avg. 64.35s

55_110_142

Part 3 -- Nbhds

83 respondents

60-Sp

90-Sp

90-Ds

120-Sp

SBGN

Metro

HOLA

yFiles

Nbhds for 60-Sp

HOLA

Avg time: 9.69s

73 -- avg. 9.94s

0_2_3_5_6_8

7 -- avg. 7.60s

2_3_5_6_8

1 -- avg. 9.45s

0_3_5_6_8

1 -- avg. 7.96s

2_5_6_8_30

1 -- avg. 7.74s

0_2_3_5_6_8_30

yFiles

Avg time: 12.24s

40 -- avg. 10.34s

0_2_5_6_8

36 -- avg. 14.37s

0_2_3_5_6_8

2 -- avg. 12.15s

2_5_6_8

1 -- avg. 11.59s

0_3_5_6_8

1 -- avg. 8.89s

2_3_5_8

1 -- avg. 17.26s

2_3_5_6_8

1 -- avg. 7.25s

0_2_5_8

1 -- avg. 16.02s

0_2_3_5_8

Nbhds for 90-Sp

HOLA

Avg time: 8.19s

82 -- avg. 8.21s

2_20_77_78_79_80

1 -- avg. 6.68s

2_20_78_79

yFiles

Avg time: 10.84s

67 -- avg. 10.80s

2_20_77_78_79_80

11 -- avg. 11.92s

20_77_78_79_80

2 -- avg. 7.45s

77_78_79_80

1 -- avg. 7.42s

20_78_79_80

1 -- avg. 14.12s

2_20_78_79_80

1 -- avg. 8.39s

2_20_77_79

Nbhds for 90-Ds

HOLA

Avg time: 13.31s

41 -- avg. 14.23s

2_6_8_9_17_19_29

11 -- avg. 13.70s

2_6_8_9_19_29

4 -- avg. 7.97s

2_9_19_29

4 -- avg. 9.41s

2_6_9_19_29

4 -- avg. 11.70s

2_8_9_17_19_29

2 -- avg. 9.55s

2_6_9_17_19_29

2 -- avg. 11.67s

2_9_17_19_29

2 -- avg. 13.45s

6_8_9_17_19_29

2 -- avg. 11.17s

2_6_9_19

2 -- avg. 7.66s

6_8_9_17_19

1 -- avg. 10.07s

2_8_9_17_19

1 -- avg. 13.72s

2_6_8_9_17_19_24_29_37

1 -- avg. 12.29s

2_6_8_9_19

1 -- avg. 53.61s

2_6_8_9_19_28_29_72

1 -- avg. 10.72s

2_6_8_9_17_19

1 -- avg. 14.35s

9_19_24_37

1 -- avg. 14.45s

2_6_8_9_17_19_29_72

1 -- avg. 7.56s

9_17_19_29

1 -- avg. 10.12s

6_9_17_19_29

yFiles

Avg time: 15.41s

46 -- avg. 16.81s

2_6_8_9_17_19_29

9 -- avg. 13.60s

2_6_8_9_17_19

5 -- avg. 11.32s

2_6_8_17_19

4 -- avg. 14.15s

2_6_8_9_19_29

4 -- avg. 15.54s

2_6_8_17_19_29

3 -- avg. 12.72s

2_6_8_19_29

2 -- avg. 13.26s

6_8_17_19

1 -- avg. 21.73s

2_6_17_19

1 -- avg. 13.81s

2_6_9_17_19_29

1 -- avg. 4.98s

6_8_19

1 -- avg. 26.65s

6_8_19_22_24_36_37

1 -- avg. 9.85s

2_6_8_17_19_36

1 -- avg. 13.16s

2_6_8_9_17_19_24_36

1 -- avg. 13.98s

2_6_8_9_17_19_29_36

1 -- avg. 16.23s

2_6_8_9_19

1 -- avg. 9.78s

2_6_17

1 -- avg. 13.21s

2_8_9_17_19_29

Nbhds for 120-Sp

HOLA

Avg time: 12.05s

72 -- avg. 12.37s

2_32_35_38_93_105_112_117

5 -- avg. 8.14s

2_35_38_93_105_112_117

2 -- avg. 14.57s

2_32_35_38_93_105_117

1 -- avg. 12.15s

35_38_93_105_112_117

1 -- avg. 10.68s

2_38_93_105_112_117

1 -- avg. 8.58s

2_32_35_93_105_112_117

1 -- avg. 8.24s

2_32_38_93_105_112_117

yFiles

Avg time: 15.55s

45 -- avg. 15.70s

2_32_35_38_93_105_112_117

16 -- avg. 13.75s

2_32_35_93_105_112_117

3 -- avg. 23.77s

1_32_35_38_93_105_112_117

2 -- avg. 10.19s

2_32_93_105_112_117

2 -- avg. 15.43s

2_32_35_38_93_105_117

2 -- avg. 13.40s

2_32_38_93_105_112_117

2 -- avg. 14.45s

2_32_35_93_105_117

1 -- avg. 11.25s

2_32_38_93_105_117

1 -- avg. 11.57s

2_38_105_112

1 -- avg. 16.05s

2_35_38_105_112_117

1 -- avg. 10.52s

32_35_38_93_105_111_112_117

1 -- avg. 8.84s

32_93_105_112_117

1 -- avg. 8.51s

32_35_93_105_117

1 -- avg. 17.47s

32_35_93_105_112_117

1 -- avg. 43.03s

32_35_38_40_80_93_105_112_117

1 -- avg. 23.55s

1_32_35_93_105_112_117

1 -- avg. 20.14s

1_2_32_35_93_105_112_117

1 -- avg. 14.83s

2_32_35_105_112_117

Nbhds for SBGN

HOLA

Avg time: 8.88s

83 -- avg. 8.88s

6_19_29_55_108_109_159

yFiles

Avg time: 13.83s

63 -- avg. 13.24s

6_19_29_55_108_109_159

10 -- avg. 16.26s

6_19_29_55_108_159

3 -- avg. 13.05s

6_19_29_55_109_159

1 -- avg. 31.49s

29_55_84_108_159

1 -- avg. 17.59s

6_19_29_55_159

1 -- avg. 9.29s

19_29_55_108_109_159

1 -- avg. 18.41s

6_29_55_108_109_159

1 -- avg. 6.51s

29_55_108_159

1 -- avg. 17.42s

6_19_29_109_159

1 -- avg. 11.51s

29_55_73_108_159

Nbhds for Metro

HOLA

Avg time: 7.26s

81 -- avg. 7.23s

59_103_122_150

1 -- avg. 6.53s

59_92_103_122_150

1 -- avg. 10.17s

103_122_150

yFiles

Avg time: 10.85s

76 -- avg. 10.79s

59_103_122_150

3 -- avg. 10.42s

59_103_122

2 -- avg. 15.57s

59_122_150

1 -- avg. 9.37s

59_103_122_150_170

1 -- avg. 9.10s

103_122_150

Part 4 -- Large Graph Pairs

83 respondents

Graph 1, HOLA, 62 votes
  • The squares are bigger and much easier to read.
  • Better logical flow
  • Distances between nodes more uniform.
  • better in terms of alignment harmony, better aesthetics.
  • boxes more defined
  • Spacing is smaller
  • Similar comment to previous. Nodes are clusterered together which is easier to graps information more easily and see the summary of what is going on. The not so good diagram is to spaced out in its layout.
  • i likethe clustering better
  • Size matters.
  • prefer shorter lines
  • No node connection that runs along the 'outside'
  • Descendents more clear
  • Same issue as before
  • Other diagram has too many long lines linking boxes, makes it hard to see what connects to what
  • this image is slightly better than the one i haven't chosen, perhaps because the nodes are larger and the use of the 'brackets' is better than straight lines (as in the first image) for indicating branches from the main node.
  • More beautiful.
  • Again -More familiar - like a family tree.
  • More links between info
  • Sections are clearer
  • larger nodes help somehow curly edges also...
  • They are both bad, hard to choose. Chosen one looks less complex. Has a hierarchical bit. Hard to say because you don't specify whether I should take into account the complexity, and I don't know they are logically equivalent, but I'll take it that both are trying to convey the same information.
  • easier to follow connections.
  • bigger blocks and less crosses
  • Nodes are larger with respect to lines. Lines are shorter. Lines not so close together
  • Looks a little later
  • Consise
  • These two are pretty close. I prefer the density and shorter edges.
  • Clear and branched.
  • better grouping and curly braces give a better sense of even distribution where used. right hand side again has wrong symmetry in terms of ratio of line length to size of squares
  • The paths are shorter and more direct
  • Simpler, easier to read
  • Structure is a bit easier to follow in this one. But not by much.
  • The right diagram can show a subordinate more clearly
  • bigger boxes
  • seems more streamlined
  • ranked into classes and subclasses
  • better spaced
  • Nodes too small in the other one. Lines difficult to distinguish where they are going (curves on the corners are quite acute and it is difficult to see the direction it links to when it joins perpendicularly to the other ones).
  • nodes are larger
  • In the worse diagram, too many links are too far apart
  • The short edges are easier to follow! Having neighbouring nodes closer together would theoretically make it easier to see and understand relationships (depending on the content presented).
  • shorter connections - the longer connection lines are too difficult to follow
  • better grouping of nodes
  • Bigger nodes

Graph 1, yFiles, 21 votes
  • Loops closed, all links feel included
  • less squares
  • better overview
  • less curls & cleaner
  • again, distances between groups of nodes make the graph easier to understand.
  • The grid arrangement makes the layout look neater.
  • The nodes are more evenly arranged.
  • simple
  • Less curved lines.
  • The hierarchies with the curved lines got a bit squashed together. The other one is better as the distance between boxes makes it easier to follow.
  • I thought the other one had too many curves.
  • circular structures tent to confuse me
  • easier to see clusters of info
  • less overlapping of lines, so its less confusing

Graph 3, HOLA, 75 votes
  • It is the best size to read, the other one is way too small.
  • bigger squares
  • More compact layout
  • first one as connections are well defined and image larger
  • Distance between nodes more uniform, easier to see where connections are.
  • this is the best because the graph seems to be better organized by node / neighbor proximity (i.e. the path between nodes can be better perceived)
  • boxes are larger, easier to see
  • Spacing is smaller
  • It has clear end-points and seperation
  • Clusters of nodes are more closely grouped and easier to read compared to diagram on left, where are are long lines travelling larger distances to get to nodes.
  • easier to read
  • better image
  • BIGGER, my poor old eyes
  • Organisation is clearer.
  • Can be seen clearly
  • Descendants more clearly visible
  • Same issue as before
  • Larger and easier to read
  • Larger boxes make it easier
  • because the image is larger and the nodes are closer together without being too crowded
  • The nodes are easier to see because they're not as far away.
  • Connected nodes are generally closer together, making their association more obvioius.
  • Clearer to see, more structured layout
  • Less long lines running through diagram
  • Branching into multiple nodes is easier to interpret
  • condense linkages
  • More structured. More hierarchical?
  • more compact layout while still having clear links. More consistent spacing between boxes.
  • Bigger blocks
  • Diagram 1 has nodes that are too small and edges that are too long. Some edges are too close. Diagram 2 is better.
  • Less long lines
  • Concise and 'cleaner' view
  • The "buses" of edges make it hard to see the neighbourhood on the bad one. Also, the good one has a much better representation of the trees on the fringes.
  • Clear and branched.
  • Less rectangular
  • same as previous ones
  • I like curly lines. Also, the larger boxes are better.
  • I can clearly see the parents and the children. I like the use of curled brackets
  • Larger and easier to decipher
  • This one is better because it's easier to follow. The long edges in the other one are hard to follow, some are much too close together.
  • The right diagram more compactness than the left one and clearly show which one is subordinate.
  • The boxes are larger and easier to see
  • is more compact
  • prefer things horizontal
  • When two nodes radiate outwards from a single node, the diagram looks better because the nodes face the same direction
  • The short edges are easier to follow.
  • more orderly and grouped
  • More compact
  • The nodes were closer together so it was easier to see how one branch related to another.
  • less paths
  • Actually I like none of them.

Graph 3, yFiles, 8 votes
  • better overview
  • seemed to flow better
  • grouping of nodes makes the graph easier to understand.
  • feels like less data points

Graph 4, HOLA, 46 votes
  • Bigger squares makes it easier to comprehend.
  • bigger squares
  • Marginally.
  • boxes are larger
  • Spacing is smaller
  • Similar comment to previous. Nodes are clusterered together which is easier to graps information more easily and see the summary of what is going on. The not so good diagram is to spaced out in its layout.
  • less space between thing
  • Too many lines that follow the same path closely
  • Bigger blobs and lines made it easier to see but if I could enlarge the diagram the other would be better as a the blobs don't go over the lines so much
  • Descendants more clearly visible
  • can convey information more easily in a smaller space
  • This is the less terrible diagram.
  • More condensed so easier to follow
  • less 'sparse' and easier to trace paths between nodes
  • Easier to see blocks
  • Nodes are larger in diagram 2, so it is marginally better than diagram 1. Both a bit scrambled.
  • looks nicer
  • Clearer.
  • It is clear to see the lines as they are not overlapped with squares
  • distance apart of square on right too far would prefer left hand lines emanating from a single square to have a little more gap between them in some cases
  • Larger nodes, shorter lines, more evenly distributed on page.
  • Size on the screen here
  • The left one show more compactness and clearly show which one is subordinate
  • shorter lines
  • bigger boxes
  • has more branches
  • not much crossing over
  • nodes are larger
  • The links in the worse diagram are too far apart
  • info won't be as small
  • The short edges are easier to follow.
  • There is less blank space in it and the connections are shorter
  • I prefer the curved connections - both of these look equally convoluted / curled in on themselves and all have lots of long connections and connections right next to each other

Graph 4, yFiles, 37 votes
  • Easier to follow
  • Too much info in each, but first is better as groupings are clearer
  • better overview
  • Both seem confusing and messy, As i have to choose, this one seems cleaner
  • Nodes aligned to a grid
  • the links are easier to follow on this one
  • graph appears to be more coherent - different groups of nodes are discernible.
  • The nodes with neighbourhoods that contain terminal nodes are clustered together, making it neater.
  • Diagram does not look as cluttered
  • Some of the connections are clearer, and there is potential for further information to be conveyed (e.g., the entire group of elements in the top left may be related in another way). The other layout is hard to follow with some lines being hard to discern
  • Straight edges make it clearer
  • the first image is almost better, i think because it is less confusing with the lines but at the same time it looks much more like a computer's hardware or a computer log so it's not appealing to the general public.
  • The other layout is difficult to differentiate between crossings and edges running close together.
  • More systematic.
  • More spaced out. Less crossing of lines
  • Shows clearly defined regions of information (i.e. clusters of information about a certain topic, then moves on to a different group of info)
  • A bit more clearer
  • simple
  • More structured. Looks like less nodes?
  • information seems to be grouped spatially, so if there is logical grouping that is illustrated better.
  • Appears to be less overlapping lines, more structured
  • These are both pretty knotty. In this one I think the less dense one is better, because it seems to have fewer crossings, which makes the edges less likely to get confused. In the denser (bad) layout, sometimes there are crossings on top of curves, which makes it really hard.
  • Better clustering of neighbours
  • When the edges are this dense it is a little easier to understand the sparse layout.
  • Prefer the straighter lines in this one.
  • less lines everywhere
  • Clustered information, nodes look grouped and ordered, even though the links are sometimes attenuated.
  • clear layout with the connecting lines
  • clear paths

Graph 5, HOLA, 70 votes
  • The other one is too small.
  • bigger squares
  • Easier to see clustering
  • all clusters linked, image large enough to view
  • more compact
  • larger size
  • Distances more uniform
  • we have a better overview of clusters and paths between clusters
  • boxes are larger
  • Spacing is smaller
  • The lines in the other one are too long, no clear end points or seperation
  • Similar comment to previous. Nodes are clusterered together which is easier to graps information more easily and see the summary of what is going on. The not so good diagram is to spaced out in its layout.
  • less space betweenthing
  • Size Matters
  • Bigger, can see the associations better
  • Branches can be more clearly followed
  • Descendents more clear
  • Tidier, especially with the nodes connected with curved lines.
  • pretier
  • Larger boxes make it easier
  • long lines mean that diagram must be very big to see clearly
  • this is more organised and shows a logical flow, branches and and clusters
  • Similar reasons to previous layouts.
  • More familiar - like a family tree.
  • Allows for more sub-headings and links between ideas
  • Connections are less confusing
  • nodal connectivity
  • Other one is too sparse. Have to move eyes too far to trace.
  • curved lines make it appear as hierarchical information at some points.
  • Clearer
  • This diagram is very nice. I like the size and shape of the nodes. The nodes look the same size and shape and the distance between edgs is not too long.
  • Less long lines
  • Better structured layout
  • The long lines in the bad one mean there is less correlation between geometric proximity and the graph's topology. The tree representation is better in the good one, but it gets a bit confusing where there are lots of trees kind of overlapping (on the right).
  • Clearer and branched.
  • BiBigger squares
  • same as previous - grouped well on left and ratio of line size to squares on right is wrong.
  • Larger, easier to follow
  • The grouping is nicer in this one. Layout is much too sparse in the other requiring close following of the edges.
  • The left one show more compactness and clearly show which one is subordinate
  • bigger boxes
  • bigger boxes
  • Larger nodes, easier to see links and relationships. Nice curves on the lines.
  • larger
  • The end-nodes radiate outwards making them easier to trace
  • info won't be as small
  • The short edges are easier to follow!!!!
  • It just makes more sense, and looks better like when when two boxes come off one box
  • shorter connections - easier to see bigger picture
  • more simplified
  • Bigger nodes

Graph 5, yFiles, 13 votes
  • again, graph appears to be more coherent in terms of group consistency.
  • The grid layout and larger spacing makes the layout look neater.
  • The nodes are far away which makes them hard to see but they are more evenly arranged.
  • This time the node size/edge ratio is better on left but ideally one with nodes sized half-way between these two would be better still....
  • Quick to navigate for direct paths
  • Although it is tiny, this one is better as the nodes are not as clustered.
  • Much more hierarchical, i cant quite see the clusters though that i can see in the other image
  • Prefer the overall shape of the other one, but the straighter lines of this one overrode this.

Graph 7, HOLA, 71 votes
  • It is much larger so it is clearer to read.
  • bigger squares
  • Seems to have more logical layout
  • too much again, but larger font gives better chance to mentally understand grouping. Groups are clearer on unselected image but too small to view
  • curvy branches are easier to see
  • other one is too small
  • Nodes closer together
  • bigger nodes, better perception of groups of neighbors and easier to follow links
  • easier to see
  • Spacing is smaller
  • Too long lines on other, unclear
  • Less useless space in between nodes clusters.
  • less space between things
  • It is clearer, the larger the image, easier to understand
  • Bigger and fewer lines crossing or at least more separate.
  • nodes are not as far apart
  • The children of a head node is clearer- so whatever relation they have is more obvious
  • Elements are easier to follow. The other has elements too far apart that would only make sense if it were representing geographical distance
  • Larger and easier to read
  • can convey information more easily in a smaller space
  • in the one i didn't choose, there are some 'outlier' nodes which don't seem relevant, although they likely are! having them separate from the other nodes in the diagram is misleading
  • Both these diagrams are terrible. I have no idea what's going on here.
  • More compact, without compromising clarity.
  • More focused.
  • More familiar - like a family tree.
  • Easy to see where connections are
  • greater interactivity
  • much easier to see relationships between nodes and read/trace paths between them
  • Other one is too sparse
  • curved lines show hierarchy. spacing is more consistent. easier to follow connections.
  • Clearer
  • Nodes are too small. Lines between them seem very long and too close together.
  • Shorter Lines
  • Layout is better structured
  • The long lines and buses where some arcs come off in the middle are just bad.
  • Clear and branched.
  • Bigger squares, shorter lines.
  • same reasons again
  • This one has larger boxes and looks more compact, making it look less complicated.
  • Lots of clusters, I can see that being valuable to see and understand
  • Larger, easier to read
  • This one is easier to to follow. The sparse layout requires more work to see where different structures link to each other.
  • The right diagram show more compactness than the left one
  • bigger overall, shorter lines
  • has more branches and is not all squished in together as opposed to the other one
  • more evenly spaced
  • The nodes are bigger. They are still grouped even though they are larger. Too much empty space in the other one.
  • As with the previous diagrams, in the worse diagram the links are too far apart
  • both horrible! but better not to have such long distances
  • The short edges are easier to follow!
  • the lines were shorter and the boxes were more grouped together
  • again nodes closer together, no huge long arms that you have to follow for ages

Graph 7, yFiles, 12 votes
  • Both are too busy and have too many entry points, but this one is cleaners with straight lines
  • grouping of nodes make graph easier to understand.
  • Terminal nodes are clustered together, making it look neater.
  • Straight edges make it easier
  • More structured
  • I didn't like the curves in the other one.
  • look more organised
  • Less cluttered
  • It is less scattered
  • appears more structured

Graph 8, HOLA, 78 votes
  • The squares are bigger which makes the paths easier to understand.
  • bigger squares
  • Better symmetry. It really depends on the use of the diagram.
  • Clearer from visibility viewpoint, lines appear to head in logical direction
  • other one is too small
  • more linear, makes it easier to follow
  • Easier to see the links. However, the choice would be dependent on what sort of information is supposed to be conveyed: if you are supposed to get an idea of distance, I would have chosen the other.
  • better because fewer links to follow and better perception of neighbor groups
  • Easier to see
  • Spacing is smaller
  • They both make me cry
  • less space between things
  • Insanely small, therefore useful for the general public
  • prefer shorter lines also bigger
  • Shorter connections make the layout less confusing.
  • less circular routes
  • Looks more logical The other has nodes too far apart that make it difficult to track the line across the entire distance
  • No grid lines make it easier
  • Squares in other diagram are confusing
  • both are unappealing and look like they are poorly designed, and i wouldn't be bothered deciphering their content. if the designer's aren't putting time into the layout then i'm not putting time into understanding the material.
  • The nodes are more evenly arranged but they are very difficult to see, but not as difficult as the other diagram.
  • Edges are very long and close together, making them difficult to visually follow.
  • Less distance between nodes. Less confusion/potential error
  • Less confusing
  • Diagram is larger and nodes are more evenly spaces. Lines are too long in the other one.
  • could not see other clearly
  • the one I didn't choose is way too fine and small and difficult to read...too sparse - i.e. too much effort to trace paths when lines are so fine...
  • Other one is too sparse.
  • can't see the boxes on the alternative.
  • clearer. less long lines
  • The first diagram hurts my eyes to look at. Definitely not suitable for short sighted people. The edges fade into the background and the nodes look like small glowing lights. The second diagram is better, but not my ideal diagram. At least everything is orderly and well spaced.
  • Shorter lines
  • No overlapping lines, more structured
  • The bad layout is an absolute disaster. The good layout has almost uniform edge length, which is great.
  • Much clearer.
  • Less lines close together
  • As there is a larger amount of squares it is easier to follow a path rather than trying to follow a thin line.
  • better use of space and relationships
  • Looks cool. Other one is too line-y
  • The other image is far too small , I can barely see the lines. This one seems to be nodes only?
  • More compact
  • Larger, more legible
  • It's very easy to get lost in the other one with the long edges. Edges probably shouldn't be so close together.
  • More symmetry
  • structure appears clearer
  • The left one show more compactness
  • less long lines
  • clearer
  • properly aligned, which makes it better to understand.
  • Cant see the lines on the other.
  • It seems like the other one is needlessly complicated.
  • nodes are larger
  • The better diagram has a more sequential flow to it
  • More spread out
  • All the edges are short and easy to follow with your eyes.
  • i like the consecutive order of the boxes, i don't know if I'm making any sense
  • The information looks like it proceeds linearly, like one piece follows the other - and the straight armed branches are clearer than what looks like loops.
  • less complicated
  • Bigger nodes

Graph 8, yFiles, 5 votes
  • more concentrated markings
  • groups of nodes are separated from others via long distances. provided these groups contain nodes with related information, this is a good way of displaying the graph.
  • Items have been grouped, implying they have something important in common- so more information is conveyed
  • First one has less nodes, makes it easier to follow